Two numbers,
pulling apart
Four cases where two official series seem to disagree. Each pair is shown as two charts — one unit each, both starting the same year — with what each side measures and why both can be true at once. No verdicts: the gap is the story.
all numbers computed from the underlying official series · refreshed nightly
More diplomas, lower test scores
Is education producing more, or less?
Tertiary attainment · 2003: 27.5 → 2025: 46.1 %
J3 Tertiary attainment (25–64) — full page, source & caveats →
PISA mathematics · 2003: 538 → 2022: 493
PISA reading · 2003: 513 → 2022: 459
The left chart counts credentials in the adult population; the right one tests what 15-year-olds can actually do. Both are official, and over the same two decades they have moved in opposite directions.
Both can be true at once. Attainment rises as older generations are replaced by younger ones who studied longer — it says nothing about what was learned. PISA measures the skills of one age group at one moment — it says nothing about degrees. Neither answers “is education getting better?” on its own; the distance between them is why this site shows both.
Longer lives, not always healthier ones
Are the extra years good years?
Life expectancy, women · 1981: 79.3 → 2025: 83.5 years
Life expectancy, men · 1981: 72.7 → 2025: 80.7 years
Healthy years, women · 1981: 62.4 → 2025: 62.0 years
Healthy years, men · 1981: 59.9 → 2025: 64.1 years
Both charts start from the same mortality tables. The right one adds a health survey: the years a newborn can expect to live in good or very good self-reported health. Lives have lengthened for four decades; the healthy share of those lives has not moved in step, and the difference between the two charts — the expected years in poorer health — has not shrunk as lives lengthened.
“Good health” here is what people say about themselves, not a clinical measure, and this CBS variant is not comparable in level with the EU’s activity-limitation measure. Directions are meaningful; precise levels are softer than they look.
The economy per head vs the median household
If the economy grows, who feels it?
Real GDP per capita · 2011: 44,980 → 2025: 51,400 € (2020 prices)
Real median income · 2011: 24.3 → 2024: 28.0 k€ (2015 prices)
The left chart is everything the economy produces, divided by everyone — including retained company profits, depreciation and taxes that never land in a household budget. The right is what the household in the middle actually has to spend, after taxes and transfers, adjusted for inflation. An average can rise while the middle moves less — and neither number says anything about who got what.
One honesty note: the two lines are inflation-adjusted with different price indices (GDP with its own deflator, income with consumer prices), and the income series only starts in 2011. Compare directions, not precise gaps.
House prices vs house building
Did prices rise because building fell?
Real house price index · 1996: 51 → 2025: 119 (2020 = 100)
⌇ 2012: No data 2012–2014: source table change (permit-based → BAG); the gap is real and is not interpolated
Completed, old register · 1996: 88,934 → 2011: 57,703
Completed, BAG register · 2015: 48,916 → 2025: 69,189
Prices on the left are already inflation-adjusted: this is how much homes rose over and above everything else. The right chart counts homes actually added to the stock each year. Completions are one lever behind prices, not the whole story — prices are also pushed by interest rates, incomes, household growth and the existing stock. This pairing shows the two numbers most often argued about; it is not a causal claim in either direction.
The completions series changes register in 2012–2014 and the gap is shown, never interpolated — the two segments are close but not perfectly comparable.
Reading rule: two separate charts, one unit each — never a shared axis. Both charts in a pair start the same year, so neither line looks flatter merely because it is longer.